Examples of opennlp.model.EvalParameters

opennlp.model.EvalParameters
This class encapsulates the varibales used in producing probabilities from a model and facilitaes passing these variables to the eval method.

    
    // The model does need the correction constant and the correction feature. The correction constant
    // is only needed during training, and the correction feature is not necessary.
    // For compatibility reasons the model contains form now on a correction constant of 1, 
    // and a correction param 0.
    evalParams = new EvalParameters(params,0,1,numOutcomes);
    int[] activeOutcomes = new int[numOutcomes];
    int[] outcomePattern;
    int[] allOutcomesPattern= new int[numOutcomes];
    for (int oi = 0; oi < numOutcomes; oi++) {
      allOutcomesPattern[oi] = oi;

View Full Code Here

      params[pi] = new MutableContext(allOutcomesPattern,new double[numOutcomes]);
      for (int aoi=0;aoi<numOutcomes;aoi++)
        params[pi].setParameter(aoi, 0.0);
    }


    EvalParameters evalParams = new EvalParameters(params,numOutcomes);
  
    /** Stores the sum of parameter values of each predicate over many iterations. */
    MutableContext[] summedParams = new MutableContext[numPreds];
    if (useAverage) {
      for (int pi = 0; pi < numPreds; pi++) {

View Full Code Here

      params[pi] = new MutableContext(allOutcomesPattern,new double[numOutcomes]);
      for (int aoi=0;aoi<numOutcomes;aoi++)
        params[pi].setParameter(aoi, 0.0);
    }


    EvalParameters evalParams = new EvalParameters(params,numOutcomes);
  
    /** Stores the sum of parameter values of each predicate over many iterations. */
    MutableContext[] summedParams = new MutableContext[numPreds];
    if (useAverage) {
      for (int pi = 0; pi < numPreds; pi++) {

View Full Code Here

    
    // The model does need the correction constant and the correction feature. The correction constant
    // is only needed during training, and the correction feature is not necessary.
    // For compatibility reasons the model contains form now on a correction constant of 1, 
    // and a correction param 0.
    evalParams = new EvalParameters(params,0,1,numOutcomes);
    int[] activeOutcomes = new int[numOutcomes];
    int[] outcomePattern;
    int[] allOutcomesPattern= new int[numOutcomes];
    for (int oi = 0; oi < numOutcomes; oi++) {
      allOutcomesPattern[oi] = oi;

View Full Code Here

      for (int oi = 0; oi < outcomePattern.length; oi++) {
        alpha[oi] = parameters[ci + (outcomePattern[oi] * monitor.getPredLabels().length)];
      }
      params[ci] = new Context(outcomePattern, alpha);
    }
    this.evalParams = new EvalParameters(params, monitor.getOutcomeLabels().length);
    this.prior = new UniformPrior();
    this.modelType = ModelType.MaxentQn;
    
    this.parameters = parameters;
  }

View Full Code Here

    display("\t  Number of Predicates: " + numPreds + "\n");
    


    params = new MutableContext[numPreds];
    if (useAverage) averageParams = new MutableContext[numPreds];
    evalParams = new EvalParameters(params,numOutcomes);
    
    int[] allOutcomesPattern= new int[numOutcomes];
    for (int oi = 0; oi < numOutcomes; oi++) {
      allOutcomesPattern[oi] = oi;
    }

View Full Code Here

    
    // The model does need the correction constant and the correction feature. The correction constant
    // is only needed during training, and the correction feature is not necessary.
    // For compatibility reasons the model contains form now on a correction constant of 1, 
    // and a correction param 0.
    evalParams = new EvalParameters(params,0,1,numOutcomes);
    int[] activeOutcomes = new int[numOutcomes];
    int[] outcomePattern;
    int[] allOutcomesPattern= new int[numOutcomes];
    for (int oi = 0; oi < numOutcomes; oi++) {
      allOutcomesPattern[oi] = oi;

View Full Code Here

TOP

Related Classes of opennlp.model.EvalParameters

opennlp.maxent.GISTrainer

opennlp.maxent.quasinewton.QNModel

opennlp.perceptron.PerceptronTrainer

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.